Here’s a math problem:
Your company sells organic baby clothes, so you need some parent influencers to get the word out. You hire a company that will charge you $2,500 to get influential bloggers to tweet about you three times each over the course of three days. After much outreach, the company secures four influential bloggers in the healthy / organic family space to tweet about you.
The bloggers are Kelly, Rita, Tania, and Jane.
Kelly has 35,122 followers
Rita has 85,621 followers
Tania has 15,009 followers
Jane has 2,559 followers
For the sake of brevity, we’re going to look at the success of Kelly’s tweets only.
Over the course of three days, the same six people retweet all of Kelly’s tweets about you.
Person one has 3,500 followers
Person two has 876 followers
Person three has 12,380 followers
Person four has 256 followers
Person five has 1,222 followers
Person six has 65 followers.
How many times did Kelly’s tweets show up on people’s Twitter timelines?
(35,122 x 3) + (3,500 x 3) + (876 x 3) + (12,380 x 3) + (256 x 3) + (1,222 x 3) + (65 x 3) = 160,263
Wow! Kelly’s tweets had 160,263 timeline impressions. That’s amazing. Her tweets reached more than 53,000 individuals! If you paid the company $2,500 for these four influencers, then let’s say Kelly’s cost was 1/4 of that, $625. Based on Kelly alone, your social media campaign was a rousing success. You paid less than one cent (~$.004) per impression. A steal!
I wish it were this easy. I wish it were simple multiplication. I wish we could pay an influencer $625 and have our tweets seen 160,263 times, after said well-respected social media hero tweets about it. Long live the influencer! The cheapest and most effective advertiser there is. Finally, marketing made easy.
I’m so sorry, but no. It’s wrong. It’s totally and completely wrong.
Recently I’ve come across metrics like this. Either marketers aren’t clear on how to properly measure social media reach and impact, or they don’t want to tell their clients the truth.
Here are four crucial metrics they’re not telling you:
One: the lifespan of a tweet
Two: the number of inactive Twitter users who follow these people
Three: the overlap in Twitter followers of people in the same influencer space
Four: the real cost per conversion
The Lifespan of a Tweet
While the true life of a tweet depends on the number of people each follower follows, how often his/her followers are online, and the time of day the tweet is published, we estimate that the lifespan of a tweet is somewhere between 10 and 30 minutes. This means that there is a very small window during which your tweet is visible.
If you’re smart about hashtags, if someone slow to the game retweets you later, or if any other number of fortuitous things happen, your tweet’s life may be longer than 30 minutes. We’ve seen people find old tweets somewhere in the depths of Twitter and breathe a little life into them when they’re relevant again. However, this is the exception and not the rule.
The Number of Inactive Twitter Users
Moz has a tool, Followerwonk, that will show you the recency of a Twitterer’s followers’ tweets. If you look at the chart below, you can see that more than 50% of this person’s robust follower base has either never tweeted, or hasn’t tweeted in the past seven days or more. While some of these people might be lurkers (meaning they look at Twitter without posting), many of them are probably not on Twitter at all. In fact, only 3,267 were likely to have been online during the 24 hours the tweet happened, and of those, only 1,392 were online in the hour of the tweet. Now remember, the life of a tweet is about 10 to 30 minutes. So how many of these people would’ve actually seen the tweet at all? Certainly not the more than 14,000 followers this account technically has.
Overlapping Twitter Followers
Hooray for Followerwonk. Another tool they have allows you to compare which Twitter followers overlap. I just searched for “popular mommy bloggers” on Google and grabbed two names that came up. I put their Twitter account info into Followerwonk, and I can see that almost 9% of their followers overlap:
Here’s the thing: that’s not necessarily bad. If someone follows both of these influential moms, and they both tweet about your product, then that person might trust you even more than if they heard about you from one of those people only. However, most social marketing companies aren’t going to take into account the overlap and will count the same person as two different people. In this respect, your marketing company is potentially both overselling themselves and selling themselves short at the same time.
If you happen to follow all four mommy bloggers above, you’re still just one person; you’re not four people. So when a marketing company reports you as four different people to inflate the number of “unique impressions,” they’re not telling the truth. However, if four of your favorite mommy bloggers are all tweeting about this awesome product, you might be more likely to buy it than four individuals!
Your marketing agency owes it to you to let you know who is converting based on this data. If 9% of the mommy bloggers’ followers overlap, we need to know what percentage of that overlap converts, and how it compares to the non overlapping followers. If that 9% converts at a higher rate than the other chunk, then you need to seek out more relationships that take into account this overlap. You can create Twitter ads that target the overlap, you can invest in another influencer campaign that focuses on that overlap.
However, if that 9% converts at the same or a lower rate, then you may want to expand so that it’s not about the overlap, but about truly unique impressions.
The Real Cost per Conversion
Okay, so it’s hard to know the real cost per conversion. There are a lot of factors that go into a conversion, and even though the web is full of robust and beautiful data points, human beings still have–and deserve–some privacy. A mommy blogger might tweet about your product one day. Three months later one of her followers might see your product on the shelf of her local bookstore and it rings a bell, even though she never clicked on or interacted with that tweet. Then she might go home, Google you, then click on your Adwords ad and buy your product.
You know what your data would show? It would show that this person Googled “YOUR PRODUCT,” clicked on an ad you created for your brand, then purchased. That would show up as a branded ad purchase. But do you know what actually drove that conversion? The mommy blogger. You cannot track that. It might look like you spent $5 on the click that converted, but in reality the cost is much more complex. Some of it goes to the cost of the mommy blogger. Some of it goes to your Adwords ad.*
While you won’t know every single point of interaction this person had with your brand, it is important to note that each of your marketing actions has the potential to shape this final conversion. For fun, you could try to track down a few of your buyers online, just to see if they might have been influenced by one of your influencers, but you’d have to do some minor internet stalking for this to work right, and, again, people deserve some privacy.
*(In some cases you’ll be able to track conversions made as a direct result of the influencer’s tweet, especially if you’re using the proper UTM codes on all those links pointing to your website).
This doesn’t sound all that bad, right? Even if I’m not getting hundreds of thousands of impressions, I know that my product and brand are reaching some people. Everyone trusts an influencer! Right? Right?
Not Everyone Trusts an Influencer
When I first started doing SEO, the influencer concept was nascent. Google’s algorithm was just beginning to catch on to fake link building, and ethical SEOs began promoting tons of fun ideas like sponsorships, guest blogs, new types of content creation, and getting involved with influencers.
And, as from all good things, out sprang a cottage industry of influencers, aka brand ambassadors. While the title of influencer was initially bestowed on bloggers and social media profiles who were doing something interesting, people now set out to be influencers rather than become them. In the past they might tweet about your product because your product was awesome, and they actually liked it. But now, they tweet about your product because you pay them to. And their tweets must show they’re being paid to do so. And their followers know it.
Some agencies that focus on influencer marketing have teamed up with these professional influencers and have basically created a network of advertisers. In fact, you’ll see an #ad tag on a tweet, and the people retweeting that tweet are often other mommy blogger influencers, so it’s basically this weird internal echo chamber where a group of people is getting paid to retweet your ad to each other.
In other words, while influencer marketing was initially seen as a way to get in front of a community of people via an authentic, almost folksy authority on a subject (parenting, cooking), now it’s just another ad that competes in an already tight space.
With that said, there are always exceptions.
What to Ask Your Influencer Marketing Agency
An influencer marketing agency is making a pitch, and it sounds pretty good. But before you sign on, ask to see some examples of their past work. Who are the people in their influencer network? Look through their tweets and see who’s interacting with the #ads. Who’s retweeting? Who’s responding? Is is the same overlapping circles? What other kind of material do these people post?
If the agency doesn’t want to provide you with this, then they may have something to hide. If that’s the case, then consider spending that money somewhere else. Facebook or Twitter ads, maybe? Google Adwords? Or find an influencer marketing agency that is more transparent and willing to give you the actual numbers, not just what sounds good.
More than once I have seen influencer marketing agencies give completely incorrect numbers, saying your tweet had 14,000 impressions when it actually only had 300. They either don’t know they’re doing the math wrong, or they do know and are doing it on purpose. Either way, you can avoid this by asking for past details and analyzing the data yourself.
As always, if you have any questions feel free to reach out to me, and I’m happy to take a quick glance at a proposal or report and see if I think the numbers are legit. firstname.lastname@example.org